The overall goal of this program project is to identify and evaluate novel natural product cancer chemopreventive agents. In the discovery process, activity-guided fractionation is employed, utilizing short-term bioassays as a monitor. The first goal of this project is to provide bioassay support of physiological relevance. As validate previously, the current test battery is comprised of assays designed to monitor inhibition of carcinogenesis at the stages of initiation (antioxidant activity, antimutagenic activity, induction of quinone reductase in cell culture), promotion (inhibition of phorbol ester-induced ornithine decarboxylase activity in cell culture, inhibition of cyclooxygenase activity0, and progression (induction of cell differentiation, anti-estrogenic activity). Test materials and assay procedures actually used for specific fractionations are keyed on secondary discriminations provided by another project. Once lead materials are structurally characterized, additional studies are performed in this project to more fully characterize the potential of the agent to succeed in clinical intervention trials. With semi-preparative quantities of active leads (0.05-1 g; supplied by projects 1 or 5), preliminary mechanistic evaluations will be performed. In addition to in vitro methodologies, this will involve short-term animal studies designed to assess effects on biomarkers of carcinogenesis (e.g., induction of phase II enzymes; inhibition of phorbol ester-induced ornithine decarboxylase activity). In addition, with relatively small quantities of test materials, inhibition of tumorigenesis will be determined in the two-stage mouse skin model. Finally, based on the overall profile of structure, activity, natural abundance, etc., the most promising test materials will be evaluated for chemopreventive activity in full-term models of carcinogenesis with laboratory animals. The specific animal and administration protocol will be based on the overall profile of the agent. The full-term tumorigenesis systems to be employed include the two- stage mouse skin model, the carcinogen (DMBA or NMU)-induced rat mammary model, the PIN (prosthetic intraepithelial neoplasia (prostate model with male Noble rats, and the benzo(a)pyrene-induced lung tumor model in Strain A mice. Complete data sets will be prepared and transmitted to organizations capable of promoting more advanced stages of development, such as Monsanto Life Sciences and the National Cancer Institute.